Saturday, November 17, 2018

What is the complement of blue?

Chris says: "James your explanation of the Yurmby wheel in your book Color and Light changed my life. It's such a great way to make cohesion in the palette as well as knowing how to move a color. What always gets me though is how do you make grey from yellow and blue, that part of my mental wiring is the last hold out."

Chris, you're right to wonder about that. If you mix yellow and blue paint, you get green, not gray. However, in the machinery of the eye, a royal blue is opposite to yellow (not orange). You can test this for yourself by looking at a blue square for 30 seconds, and then let your eyes move to the white area below it. 

 What color did you see in the white box? Most people see yellow, not orange.

It turns out that complements are different in additive mixtures (the realms of visual perception, theater lighting, digital art, and camera sensors) than they are in subtractive mixtures (pigments).

J.M.W. Turner, Norham Castle, Sunrise
The argument for thinking in terms of the additive (blue-yellow) complements when painting is that you might as well optimize your final image to the behavior of the eye rather than to the behavior of the pigments, because that's what matters to the viewer's experience.
Read More: 
Previously on this blog: The Color Wheel, Part 3: Complements and Afterimages
On the excellent website: HueValueChroma: Additive Complementaries
On the excellent website: HueValueChroma: Subtractive Complementaries
Color and Light


Glenn Tait said...

Michal Bach has a great online app that enables you to easily see visual compliments using Hinton's Lilac Chaser.
This is why I prefer Mussel's colour model and complementary relationships because they are based on how the eye sees colour.

A Colonel of Truth said...

And one reason, I suppose, the great Sergei Bongart kept Indian Yellow close at hand - to exploit blue.

Terrace said...

What does it mean that I saw a pale peach color?

Jim Williams said...

I saw a pale yellow square with a pale green fringe.

CBreier said...

Color theory makes perfect sense when you're mixing light on a computer screen. Artists paints? Not so much.. Visual complements aren't the same as mixing complements. Van Dyke Red is a mixing complement of Phthalo Blue. It makes black when you mix them. Unfortunately I've only been able to fin it in the Liquitex heavy body line of acrylics. You can also mix Phthalo Blue with Pyrrole Orange or Vat Orange to make black.

Joel Fletcher said...

I find it interesting that modern color theory, such as the Yurmby Color Wheel, has replaced blue with cyan. So that Cyan, Red, and Yellow are the pure pigment colors, as we observe in printer inks for example. However, subjectively Cyan looks to be a mixture, whereas Blue seems pure. It may not be scientific, but the traditional triad of Blue, Red and Yellow speak truth to me as the only unadulterated pure colors.

James Gurney said...

Joel, It has taken me a while to think of blue and cyan as distinct basic hues, not just two kinds of blues. You're right that cyan is one of the printers' primaries, but blue is a primary color of most systems of lighting. So as you say, in the world of printers' inks, blue can be regarded as a secondary, while in theater lighting, cyan is a secondary. From the perspective of the Yurmby wheel, which spans subtractive and additive systems, all six colors are regarded as co-equal primaries.

Unknown said...

I find it interesting that, like Terrace I saw a pale peach color (after reading the 'most people see yellow' line), something close to rgb hex fff4ea which I definitely would not call yellow. I wonder if this difference in perception has something to do with my dislike of highly saturated yellow in large amounts and my enjoyment of deep blues and other cool jewel tones. It is so amazing how different one person's perception can be from another's.

Piya said...


Speaking of color, have you ever made a Cornell box? I know Nathan Fowkes uses it to teach his students but I'm more under the impression students paint it from their imagination. It seems it's more used to test the accuracy of rendering software. I'm surprised it's not used in art schools to teach color and light.



Unknown said...

Great stuff. Thanks so much for the post. Vision is extraordinary, Turner is amazing.

Yes, I agree Terrace and Anna: I'm seeing a very light, creamy very, very warm yellow-ish (?). Extremely light though. The blue above is a slightly violet Blue. If we remove the redish-ness, try a blue more robin's egg, then we'd see a very light much warmer yellow that we may think of as more towards orange. Makes sense with the CMY framework. The Photoshop comp model goes through a mid value to get a visual comp colour - so PS comps appear much richer, though very similar hue to the after images we see which are more like inverted in value, that is a rich dark in reality gives very, very light after image and PS a comp gives a much more saturated version as a visual comp.

Chris, just for fun, as to getting gray from "B + Y": I have played with Golden's Primrose Yellow and nailed a neutral gray with Ultramarine Violet. That's a very "acidic" green-ish yellow, so the UV makes sense. A slightly warmer yellow? A slightly less violet blue to get the gray. But it's very tricky as the slightest addition of paint creates a strong colour jump unlike anywhere else across possible comp mixing pairs, in my experience. A delicate hand is needed, start with a pool/blob of yellow - adding just the littlest dot of the "blue" and judge by eye where the mix is going. Does it pass by gray? What is the bias? The comp of that would be the direction I'd need to shift in changing my blue mix. Fun stuff. Not practical for making a gray, but a fun mixing exploration.

Thanks, Glenn - that's a great app.

As always insightful and generous posts, and very interesting comment section, too!

Magne Myhren said...

Hello. I'm using your book Color and Light when I'm teaching my students at The University of Agder (Southern Norway) , Visual arts. Its an eye-opener to them. Its totally new to them and everybody in every school level seems to rely totally on Johannes Itten. I would like to dig into this matter in an academic way, but its hard to find academics who have written any usable text about it. Personally I find blue and yellow a far stronger compliment to each other than blue and orange, or violet and yellow. How, however could we merge interesting findings from Johannes Ittens Color theory of contrasting colors with this new view (Yurmby)? It could be possible to combine your findings (as an artist) with Goethes (as a philosopher and scientist), but what about the contrast-theories...?
Thank you for inspirational and eye-opening art, blog etc.!
Magne Myhren

James Gurney said...

Magne, a good website that talks about additive vs. subtractive complements is HueValueChroma. You can find one page on the topic here: